Вы находитесь на странице: 1из 10

This article appeared in a journal published by Elsevier.

The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
Author's personal copy

Engineering Structures 32 (2010) 2950–2958

Contents lists available at ScienceDirect

Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct

Performance-based plastic design method for buckling-restrained braced frames

Dipti R. Sahoo a , Shih-Ho Chao b,∗
Department of Civil Engineering, Indian Institute of Technology Bhubaneswar, Orissa 750013, India
Department of Civil Engineering, University of Texas at Arlington, TX-76019, USA

article info abstract

Article history: This paper presents a performance-based plastic design (PBPD) methodology for the design of buckling-
Received 21 March 2010 restrained braced frames (BRBFs). The design base shear is obtained based on energy–work balance using
Received in revised form pre-selected target drift and yield mechanism. Three low-to-medium rise BRBFs (3-story, 6-story and 9-
25 May 2010
story) were designed by the proposed methodology and their seismic performance was evaluated through
Accepted 25 May 2010
Available online 17 July 2010
extensive nonlinear time-history analyses using forty ground motions representing the DBE and the MCE
hazard levels. Both isotropic and kinematic hardening characteristics of buckling-restrained braces were
considered in the modeling of their force–deformation behaviors. All BRBFs considered in this study
Buckling-restrained braced frames reached the intended performance objectives in terms of yield mechanisms and target drift levels. Since
Performance-based design PBPD is a direct design method, no iterations were carried out to achieve the performance objectives of
Seismic BRBFs.
Yield mechanism © 2010 Elsevier Ltd. All rights reserved.
Energy–work balance

1. Introduction frames (CBFs), such as, sudden degradation in strength and stiff-
ness, reduced energy dissipation capacity, and limited ductility. Be-
Buckling-restrained braced frames (BRBFs) are emerging sys- cause of the nearly symmetrical behavior of BRBs in tension and
tems used as primary lateral load resisting systems for buildings in compression resulting in much smaller unbalanced vertical brace
high seismic areas. The main characteristics of buckling-restrained forces, BRBFs also require smaller beam sections as compared to
braces (BRBs) are enhanced energy dissipation potential, excellent conventional CBFs with chevron bracing configurations [7].
ductility, and nearly symmetrical hysteretic response in tension It is expected that BRBFs will experience large inelastic
and compression. Different types of BRBs have been developed deformation when subjected to major earthquake ground motions.
and tested in United States and elsewhere in recent years [1]. A However, most current seismic design methods are still based on
typical BRB consists of a yielding steel core encased in a mortar- the elastic analysis approach and use indirect ways to account for
filled steel hollow section to restrain buckling, non-yielding and inelastic behavior. As such, the current performance-based design
buckling-restrained transition segments, and non-yielding and un- methodology relies heavily on an iterative ‘‘Assess Performance’’,
restrained end zones (Fig. 1). The length of the buckling-restrained ‘‘Revision Design, and ‘‘Assess Performance’’ process to reach a
(core) segment of BRB is about 60%–70% of the total length between design capable of achieving the intended performance [8]. On
work points (e.g [2,3]). Axial forces in BRBs are primarily resisted the other hand, the proposed design methodology addresses the
need for developing a systematic design approach that results in
by steel cores which are laterally braced continuously by the sur-
predictable and targeted seismic performance of structures under
rounding mortar and steel encasement to avoid their buckling un-
stated levels of seismic hazards. This in turn minimizes, if not
der compressive loads. This allows the steel core to yield in tension
totally eliminates, the assessment and redesign tasks as required
and compression, thereby significantly increasing the energy dissi-
by current code provisions.
pation capacities of BRBs as compared to conventional steel braces.
A more comprehensive background on BRBs can be found else-
2. Objectives and scope
where [4]. Recent analytical and experimental studies [5,6] have
shown that BRBFs can be used to overcome several potential prob-
Since Housner [9] applied the concept of energy into seismic
lems associated with the conventional steel concentrically braced
design of structures, much effort has been made in the field
of energy-based seismic engineering. A recently developed
performance-based plastic design (PBPD) methodology is consid-
∗ Corresponding author. Tel.: +1 817 272 2550; fax: +1 817 272 2630. ered in the present study in which inelastic characteristics of struc-
E-mail address: shchao@uta.edu (S.-H. Chao). tural components are directly considered in the design to achieve
0141-0296/$ – see front matter © 2010 Elsevier Ltd. All rights reserved.
Author's personal copy

D.R. Sahoo, S.-H. Chao / Engineering Structures 32 (2010) 2950–2958 2951

In the above equation, βi represents the shear distribution

factor at ith level and can be expressed as βi = ( j=i wj hj /
wn hn )0.75T ; βi+1 represents the shear distribution factor at (i +
1)th level; wj is the seismic weight at jth level; hi is the height at
ith level from the base; hj is the height at jth level from the base; hn
is the height of roof level from the base; wn is the seismic weight
at roof level. The energy modification factor (γ ) given in Eq. (1) can
be related to structural ductility factor (µs = 1u /1y ) and ductility
reduction factor, Rµ , by the following expression [14]:
Fig. 1. Components of a typical BRB.

γ = (2µs − 1)/R2µ . (3)

The value of Rµ for a structural system can be determined by
using the Rµ − µs − T relationship, such as an inelastic spectrum
proposed by Newmark and Hall [15], as shown in Fig. 3(a). Several
structural systems exhibit significant reduction in strength and
stiffness resulting unstable and ‘‘pinched’’ hysteresis response at
the higher inelastic deformation levels. This reduction in energy
dissipation capacity can also be accounted in the design by using
an energy reduction factor (η = A1 /A2 ) as shown in Fig. 3(b). Since
BRBFs exhibit full and stable hysteretic response, the value of η can
be assumed as unity in Eq. (1).
Δ Δ 3.2. Step-by-step design procedure

Fig. 2. Energy–work balance concept.

A step-by-step PBPD procedure of a typical BRBF system is
summarized as follows:
the desired performance objectives of BRBFs. This design method- 1. Select intended target drift ratio, θu , and desired yield
ology has already been successfully applied to various steel struc-
mechanism for expected hazard level. Estimate fundamental
tural framing systems [10–12]. This study presents an application
period (T ) for the system from the mass and stiffness properties.
of PBPD method to low-to-medium rise BRBFs with both chevron
Empirical formulae based on codes (e.g. ASCE 7-05 [16]) can
and X-shaped bracing configurations. The robustness of proposed
design methodology is verified through a series of nonlinear time- also be used to estimate the expected fundamental period of
history analysis using PERFORM-3D [13] for both design basis the system.
earthquake (DBE, 10% probability of exceedance in 50 years, re- 2. Compute the plastic drift ratio (θp ) by deducting the yield
turn period = 475 years) and maximum considered earthquake drift ratio (θy ) from the pre-selected target drift ratio (θu ).
(MCE, 2% probability of exceedance in 50 years, return period = The upper-bound values of yield drift ratio for BRBFs can be
2475 years) hazard level ground motions. obtained by conventional pushover analysis based on code-
based lateral force distributions. Based on extensive nonlinear
3. Performance-based plastic design (PBPD) dynamic analyses, Sahoo and Chao [17] proposed a simple
expression for yield drift ratio as a function of total height (H)
3.1. Design philosophy of BRBFs and can be expressed as follows:

The detailed underlying methodology of PBPD can be found θy (%) = 0.2 + H /155 (Unit of H is meter). (4)
elsewhere [14]. In summary, PBPD concept uses pre-selected
3. Compute the value of α from the modal properties and the
target drift and yield mechanism as design criteria. In this
plastic drift level using Eq. (2) and estimate the value of γ
method, the design base shear is computed using an energy–work
balance concept where the energy needed to push an equivalent using Eq. (3). Determine the design base shear ratio (V /W ) of
elastic–plastic single-degree-of-freedom system up to the target structure using Eq. (1) and distribute the lateral load at various
drift level is calculated as a fraction of elastic input energy obtained story levels based on a lateral load distribution proposed
from the selected elastic design spectra (Fig. 2). The design base by Chao et al. [18] that accounts the inelastic behavior. The
shear for a structure can be expressed by [14]: magnitude of lateral load at ith story level was obtained as the
 q  design base shear times (βi − βi+1 ), where βi is defined earlier
V /W = −α + α 2 + 4 (γ /η) Sa2 /2 (1) and βi+1 = 0 at the roof level.
4. Determine brace sizes by resolving the computed story shear
where, V is the design base shear; W is the total seismic weight in the direction of braces for the nominal values of yield
of the structure; Sa is the spectral response acceleration obtained strengths of BRBs. The compressive yield strength of BRBs can
from code design spectrum; α is a dimensionless parameter be conservatively assumed as 10%–25% higher than their tensile
depends on fundamental period (T ), modal properties, and pre- yield strengths [19].
selected plastic drift ratio (θp ) and can be given by:
5. Determine the section sizes for beams and columns (termed
 0.75T −0.2 as non-yielding members) of BRBFs based on capacity design
 w h  philosophy for the maximum demand expected from the BRBs
8θp π 2
 n n  at the ultimate states. Check the compactness and lateral
α= (βi − βi+1 )hi  n . (2)
gT 2

bracing requirements of these members as per the AISC Seismic
wj hj
P 
Provisions [20].
Author's personal copy

2952 D.R. Sahoo, S.-H. Chao / Engineering Structures 32 (2010) 2950–2958

a b

Fig. 3. (a) Relationship between Rµ , µs , and T [15]. (b) Definition of energy reduction factor, η.

a b c

Fig. 4. Plan views of study buildings (a) 3-story (b) 6-story (c) 9-story.

4. Building models Response modification factor, R, equal to 8 was used in the

original code-compliant design of the 3- and 6-story BRBFs,
Three buildings (i.e., 3-, 6-, and 9-story) were considered for whereas R equal to 7 was used in the original code-compliant
design and evaluation of their seismic performance in this study. design of the 9-story BRBF [20]. However, the change in value of
All buildings were located on firm soil (site classification D) in Los R from 6 to 8 in the design does not change the overall seismic
Angeles. The dimensions and floor masses for both 3- and 6-story performance of BRBFs significantly [5]. For all cases, the occupancy
buildings were exactly matched with those adopted by Sabelli [5] importance factor, I, was assumed as unity for all code-compliant
for braced frame studies. However, plan layout and floor masses for design. Note that, R and I factors are not explicitly required in the
the 9-story building were adopted from Gupta and Krawinkler [21]. PBPD methodology. As expected, the design force level increases
As shown in Fig. 4, typical bay width in each direction of all with the increase in the value of I in an attempt to lower down the
buildings was 9.14 m (30 ft). The 6-story building had a total of drift and ductility demands of structures for a given level of ground
six bays with BRBs in each direction, whereas both 3- and 9-story shaking [22,23]. However, such procedure cannot be considered as
buildings had four bays with BRBs in each direction. All braced a direct design method to achieve the intended purpose, such as
bays were located at the perimeter of buildings. Except for the first damage control. The reduction of potential damage in structures
story height of 5.49 m for the 6- and 9-story buildings, typical story should better be handled by using appropriate drift limitations
height for all buildings was 3.96 m (Fig. 5). Seismic weights for directly. In that regard, the approach for calculating the design
the 3-, 6-, and 9-story buildings were 28 910 kN, 59 295 kN, and base shear in the PBPD method uses target drift as the governing
97 325 kN, respectively. The spectral acceleration values at 0.2 s parameter, which should account for occupancy importance. The
(SDS ) and 1 s (SD1 ) were 1.39 and 0.77g for both the 3- and 6-story values of base shear coefficient, Cs , were 0.174 for the 3- and 6-
BRBFs [5], whereas the corresponding values used for the 9-story story BRBFs and 0.085 for the 9-story BRBF, if designed based
BRBF were 1.11 and 0.61 g [21]. on current code provisions [16]. However, the design base shear
coefficients as per the PBPD methodology were 0.158, 0.100 and
4.1. Design base shear for BRBFs 0.054 for the 3-, 6- and 9-story BRBFs, respectively. Thus, the
PBPD methodology uses smaller design base shear as compared to
Table 1 summarizes various parameters used to compute the current codes with the maximum reduction in design base shear as
design base shear for all BRBFs used in the present study. A 43% and 36% for the 6-story and 9-story BRBFs, respectively. Since
target drift level of 1.75% was selected for all BRBFs leading to the collapse prevention rather than drift control is the governing
structural ductility factor, µs , ranging from 3.94 to 6.25. The performance criteria for structures under MCE hazard level, the
computed range of ductility was smaller than the brace ductility performance of BRBFs designed for DBE level was also investigated
capacities according to prior experimental studies [6,19]. A pinned under MCE ground motions in this study.
beam–column-brace connection, as illustrated in Fig. 5, was
used at all story levels of both 3- and 6-story BRBFs to avoid 4.2. Design of frame members
undesirable connection failures due to induced moments resulted
from unbalanced brace forces [6]. However, for the 9-story BRBF The desired yield mechanism of BRBFs requires that yielding
with X-configuration BRBs, pinned beam-to-column connections should be limited to braces and plastic hinges at the bases of first
were used at the story levels where unbalanced brace forces story columns only (Fig. 2). All frame members were designed
were present, otherwise rigid connections where unbalanced brace based on their nominal material strengths. The maximum axial
forces were absent. force demand on BRBs at the target drift level was obtained from
Author's personal copy

D.R. Sahoo, S.-H. Chao / Engineering Structures 32 (2010) 2950–2958 2953

Table 1
Design parameters used in the computation of design base shear for PBPD BRBFs.
No. Parameters 3-story 6-story 9-story Note

1 Target drift ratio, θu (%) 1.75 1.75 1.75 Pre-selected

2 Total frame height (m) 11.9 25.3 37.2 Fig. 5
3 Yield drift ratio, θy (%) 0.28 0.37 0.44 Eq. (4)
4 Fundamental period (s) 0.43a 0.77a 1.03a Ref.: ASCE 7-05 [16], Cu = 1.4
5 Inelastic drift ratio, θp (%) 1.47 1.38 1.31 (5) = (1)−(3)
6 Ductility reduction factor, Rµ 4.79 4.78 3.94 Fig. 3(a)
7 Structural ductility factor, µs 6.25 4.78 3.94 (7) = (1)/(3)
8 Energy modification factor, γ 0.51 0.38 0.44 Eq. (3)
9 Spectral acceleration, Sa 1.39 1.00 0.59
10 Base shear coefficient, V /W 0.158 0.100 0.054 V = 1170 kN (3-story); 1005 kN (6-story);
1350 kN (9-story)
Based on computer analysis, fundamental periods are 0.65 s. (3-story), 1.03 s (6-story), and 1.44 s (9-story).

Table 2
Details of BRBs and structural sections used in PBPD BRBFs.
Members Story 3-story 6-story 9-story

1st 1093 1093 1137

2nd 937 885 942
3rd 624 833 907
4th – 729 856
BRBs (Tensile yield strengths, kN) 5th – 572 789
6th – 364 703
7th – – 596
8th – – 464
9th – – 296
1st–2nd W14 × 90 W14 × 145 W14 × 233
3rd–4th W14 × 90 W14 × 145–W14 × 74 W14 × 145
Column sections
5th–6th – W14 × 74 W14 × 90
7th–9th – – W14 × 61
Beam sections – W14 × 34 W14 × 30 (1st–3rd) W16 × 40
(All floors) W14 × 26 (4th–6th) (All floors)

based on expected yield and ultimate strengths of BRBs in tension

and compression by applying material overstrength factor (Ry ),
compression adjustment factor (β ) and tension adjustment factor
(ω) to the nominal yield strength values. Since these values depend
on the type of braces used in BRBFs, two different sets of values
were considered in this study. For 3- and 6-story BRBFs, the values
of Ry , β , and ω were considered as 1.3, 1.1, and 1.4, respectively.
The corresponding values of Ry , β , and ω for 9-story BRBF were
1.1, 1.22, and 1.45, respectively [25]. For all cases, nominal yield
strength of materials used in beams and columns was 50 ksi with
the value of Ry as 1.1. Due to the introduction of pinned beam-to-
column connection (i.e., beam splice), all columns of BRBFs were
designed for axial loads only. It was later shown in nonlinear
time-history analyses that additional moments induced by column
bending have negligible effect on the overall seismic behavior.
Table 2 summarizes the details of BRBs and beams and columns at
various story levels of BRBFs. Chao et al. [18] showed that relative
story shear distribution for a braced frame using a lateral load
distribution factor of 0.75 represents an upper bound to nonlinear
dynamic analysis results and leads to more uniform deformations
of elements as well as stories over the height of the structure as
compared to a factor of 0.50. Hence, all members of BRBFs in this
Fig. 5. BRBFs and their brace-beam–column connections (a) 3-story (b) 6-story (c)
9-story. study are designed for a load distribution factor of 0.75.

the expected story shear by using a lateral force distribution 5. Seismic performance of BRBFs
proposed by Chao et al. [18]. Fig. 6 shows a step-by-step procedure
for design of frame members of BRBFs with chevron bracing 5.1. Modeling and analysis
configuration. Similar procedure with minor modifications in brace
unbalanced forces can be followed for design of BRBFs with braces The seismic performance of BRBFs was evaluated by nonlinear
in X-configuration. Brace sizes were determined by assuming analysis using a computer program PERFORM-3D [13]. Beams and
nominal yield strength of braces as 36 ksi and strength reduction columns were modeled as standard frame elements with plastic
factor of 0.9 for both tension and compression [24]. However, the hinges lumped at both ends. Both axial-moment (P-M-M) and
maximum force demands on beams and columns was computed moment–rotation hinges were assigned to all beam and column
Author's personal copy

2954 D.R. Sahoo, S.-H. Chao / Engineering Structures 32 (2010) 2950–2958

Fig. 6. Flowchart for design of BRBFs with chevron bracings (Note: φt = φc as per AISC Specifications [24]; α = Angle of inclination of brace with horizontal; Pyt and Pyc are
BRB yield strengths in tension and compression; Put and Puc are BRB ultimate strengths in tension and compression; Fh and Fv are horizontal and vertical loads on beams;
Ptransverse = Axial load on column from transverse direction; Pbeam = Axial load on column from beam gravity loading; Vstoryshear = Design story shear; Abrb = Cross-sectional
area of BRB).

elements assuming that they would carry significant axial force end segments of BRBs were assumed as 160% and 220% of the
with biaxial bending. However, since all analyses were carried area of the core segment, respectively. Similarly, the length of
out on two dimension frame models, there would be no biaxial transition and end segments were assumed as 6% and 24% of the
bending of frame members. For the 9-story BRBFs, beam-to- total length of BRB [2]. Considering the variation of cross-sectional
column connections were assumed as rigid where BRBs and gusset area along the length of brace, the effective axial stiffness of BRB
plates were present; otherwise these connections were assumed (Keff ) should consider stiffness of all segments and can be expressed
as pined. For both 3- and 6-story BRBFs, moment releases were as follows [2].
used at beam ends to simulate the presence of beam splice (Fig. 5).
(Keff ) = EAc Aj At /(Ac Aj Lt + Lc Aj At + Ac Lj At ). (5)
Addition of the beam splice can eliminate the moment-frame
action and prevent failure in the gusset regions [3,26]. All columns Elastic modulus of steel was considered as 200 GPa to compute the
were assumed to be perfectly fixed to the ground. axial stiffness of BRBs. The post-yield stiffness of BRBs in tension
P-Delta effect due to gravity loads resulted from gravity frames can be different from that in compression depending on the type
was modeled by an equivalent continuous column representing of outer casing and confining material used for lateral support to
all gravity columns associated with the frame. The magnitude of brace core [19,27]. In this study, the post-yield stiffness of core
axial load in these equivalent columns was computed from the segments in tension and compression was assumed as 3% of their
total building weight (exclusive of tributary gravity load to the initial stiffness. It should be noted that, in practice, the presence
braced frames) and the number of braced frames along a particular of gusset plates may alter the lateral stiffness of BRBFs to some
direction. Lateral stiffness and strength of these columns at each extent [3]. Both isotropic and kinematic hardening characteristics
story level represent the sum of respective values of all gravity were considered in the modeling of force–deformation response of
columns at that story level assuming their weak axis bending. BRBs. Various hardening parameters were obtained by comparing
These columns were pinned at their bases and constrained to the hysteretic response of a typical BRB from PERFORM-3D [13]
match the frame displacement at each floor level by using pin- with the component test results [19] as shown in Fig. 7(b). To
ended rigid beams. For time-history analyses, Rayleigh damping of monitor the magnitude of plastic displacements of BRBs under
2% was used in all modes of structures. various ground motions, the values of maximum ductility of BRBs
Standard BRB elements were chosen to model all braces of were assumed as 15 and 25 for DBE and MCE level analyses,
BRBFs in PERFORM-3D [13]. In general, the area of transition respectively. Similarly, the respective values of cumulative plastic
(elastic) and end zones of BRBs are larger than that of the core displacement of BRBs were fixed as 200 and 400 times their yield
(restrained yielding) segment to limit the yielding to the core displacements. These ductility values are quite conservative since
segment only. As shown in Fig. 7(a), the area of transition and component test results showed that BRBs can have maximum
Author's personal copy

D.R. Sahoo, S.-H. Chao / Engineering Structures 32 (2010) 2950–2958 2955

Fig. 7. (a) Modeling of different segments of BRBs (b) Calibration of force–deformation characteristics of BRBs in PERFORM-3D (1 kips = 4.448 kN; 1 in. = 25.4 mm).

ductility and cumulative ductility capacities of 26 [6] and fairly uniform over the building height for 9-story BRBF. The larger
1700 [19], respectively, without any degradation in their strength interstory drifts at the lower stories of 9-story BRBF indicated
and stiffness. that BRBs at these story levels dissipated relatively larger seismic
As stated earlier, the seismic response of BRBFs was evalu- energy than those at upper stories.
ated under two earthquake hazard levels, namely, design basis Fig. 8(b) also shows the interstory drift response of BRBFs
earthquake (DBE) and maximum considered earthquake (MCE). under MCE level ground motions. The maximum value of mean
Two suites of ground motions records developed by Somerville interstory drift ratio was about 3.5% for 3-story BRBF. Similarly, 6-
et al. [28] for a hypothetical site in downtown Los Angeles with a story BRBF showed the maximum value of mean interstory drift
probability of exceedance of 10% and 2% in 50 years were selected ratio of 3.83% at the second story level which reduced to 2.08%
for nonlinear time-history analysis. These acceleration time histo- at the sixth story level. The maximum value of mean interstory
ries were derived from historical recordings or from physical simu- drift ratio of 9-story BRBF was 4.3% observed at the second story
lations and were modified such that their mean response spectrum level and reduced to 1.31% at the roof level. Thus, BRBs in the
matches the 1997 NEHRP design spectrum. A total of forty records lower stories of both 6- and 9-story BRBFs were subjected to larger
were obtained from twenty ground motions in both fault-parallel drift demands under MCE level ground motions. However, since
and fault-normal orientations. The magnitude of earthquakes var- the collapse prevention rather than controlling maximum drift of
ied from 6 to 7.4 with the hypocentral distance of earthquakes var- BRBFs is the governing performance criteria for MCE level ground
ied from 1.2 to 36 km. motions, the large values of interstory drift ratios did not hinder the
applicability of BRBFs as primary seismic resisting systems since
BRBFs with the beam–column-brace configuration shown in Fig. 5
5.2. Analysis results and discussion
have exhibited excellent seismic performance and sustain only
minor yielding at story drift ratio up to 4.8% [6]. Hence, all BRBFs
Nonlinear time-history analyses were carried out to evaluate
exhibited accepted overall performance of under MCE level ground
the seismic performance of BRBFs, in terms of interstory drift ratio,
motions even though they were designed for DBE level only.
residual drift ratio, yield mechanism and displacement ductility.
Fig. 9(a) shows the residual drift ratio response of BRBFs under
Interstory (or residual) drift ratio was defined as the ratio of
DBE level ground motions. The mean values of residual drift
the interstory (or residual) displacement to the corresponding
ratio for 3-story BRBF varied from 0.45% to 0.70%. The 6-story
story height. Due to well-controlled drifts realized by the PBPD
BRBF exhibited mean residual drifts ranging from 0.36% to 0.61%.
approach, no additional design force was used in the design to
For both 3- and 6-story BRBFs, the residual drift response was
account for the P-Delta effect. A statistical analysis was carried out fairly uniform over the building height under DBE level ground
to evaluate the mean and standard deviation of drift ratios for all motions. Similarly, the 9-story BRBF showed mean values of
BRBFs. Fig. 8(a) shows the interstory drift response of all BRBFs residual drift ranging from 0.27% to 0.67%. Similar to the interstory
under DBE level ground motions. The maximum value of interstory drift distribution, the residual drift distribution in 9-story BRBF
drift ratio for 3-story BRBF was 3.0% observed for LA13 ground was not uniform over the height. As shown in Fig. 9(b), both 3-
motion. The mean and mean-plus-standard deviation values of and 6-story BRBFs under MCE level ground motions showed the
interstory drift ratios for the 3-story BRBF varied from 1.44% to mean values of residual drift ratio as about 1.7%. The maximum
1.62%, and 2.0% to 2.33%, respectively. Similarly, 6-story BRBF value for mean residual drift ratio was 2.28% for 9-story BRBF.
exhibited a maximum interstory drift ratio of 3.3% observed for It should be noted that the distribution of residual drift ratio in
LA09 ground motion. The mean and mean-plus-standard deviation 9-story BRBF under the MCE level ground motions was different
values of interstory drift ratios for the 6-story BRBF varied from from that under DBE level ground motions and the BRBs at the
1.15% to 1.73%, and 1.56% to 2.36%, respectively. The maximum lower stories dissipated significant amount of energy due to large
interstory drift ratio for 9-story BRBF was 2.82% for LA03 ground deformations under MCE level earthquakes.
motion, whereas the mean and mean-plus-standard deviation No plastic hinges were formed in beams and columns except the
values of interstory drift ratios ranging from 1.03% to 1.72%, and yielding at column bases in BRBFs under DBE level ground motions.
1.37% to 2.35%, respectively. The average values of maximum Thus, the intended yield mechanism was achieved under the DBE
interstory drift ratio for 3-, 6- and 9-story BRBFs were 1.71%, hazard level. Under MCE level earthquakes, flexural yielding of
1.73% and 1.83%. Hence, all BRBFs reached the target drift level of beams and columns at different story levels of 9-story BRBFs
1.75% under DBE level ground motions even without any design was noticed because of larger strength and deformation demand.
iteration. Since drifts were well controlled by considering inelastic However, these members did not reach their ultimate deformation
behavior directly in the design of these BRBFs, the P-Delta effect and load-carrying capacity which prevented these BRBFs from
had no appreciable influence on their overall behavior. Unlike 3- their complete collapse. Braces did not reach their maximum
and 6-story BRBFs, the distribution of interstory drift ratio was not ductility and cumulative displacement ductility limits of 15 and
Author's personal copy

2956 D.R. Sahoo, S.-H. Chao / Engineering Structures 32 (2010) 2950–2958

a b

Fig. 8. Interstory drift response of BRBFs (a) DBE (b) MCE.

200, respectively, for all BRBFs under DBE level ground motions. motions, respectively. However, this may not be an indication of
The mean values of maximum ductility demand of BRBs were 7.3, failure of BRBs because most prior isolated BRB tests [6] were
8.0, and 9.3 for 3-, 6-, and 9-story BRBFs, respectively. Similarly, the carried out under smaller maximum ductility demand (10–25) and
mean values of cumulative displacement ductility demand of BRBs high cumulative ductility demand (approximately 300–1600), as
under the DBE level ground motions were 24.5, 21.8, and 32.7 for 3- opposed to the smaller cumulative ductility demand (43.6–144.1)
, 6-, and 9-story BRBFs, respectively. BRBs in all BRBFs also did not but higher maximum ductility demand (29.4–35.6) observed in the
exceed their cumulative ductility demand of 400 under the MCE time-history analyses under MCE ground motions.
level ground motions. The mean values of maximum cumulative
displacement ductility demands were 43.6, 48.7, and 47.2 for 3-, 6. Summary and conclusions
6-, and 9-story BRBFs, respectively, under MCE level ground
motions. Similarly, the mean values of maximum ductility demand A direct design methodology, called performance-based plastic
of BRBs of 3-, 6-, and 9-story BRBFs under the MCE level ground design (PBPD), based on energy–work balance, pre-selected target
motions were 17.2, 17.6, and 16.0, respectively, which indicates drift and yield mechanism was developed in this study to achieve
that mean values of maximum ductility demand of BRBs did not predictable behavior of BRBFs by incorporating the inelastic
reach their maximum ductility limits of 25. The absolute maximum characteristics of components in the design. The PBPD design
value of maximum ductility demand of BRBs of 3-story and 6-story procedure is not too different from what is done in current practice,
BRBFs were about 32.1 and 32.3 under LA 18 and LA 16 ground yet it can be readily incorporated within the context of broader
Author's personal copy

D.R. Sahoo, S.-H. Chao / Engineering Structures 32 (2010) 2950–2958 2957

a b

Fig. 9. Residual drift response of BRBFs (a) DBE (b) MCE.

Performance-Based Earthquake Engineering (PBEE) framework. 1. BRBFs designed as per PBPD methodology can successfully
It does differ from the way PBEE is practiced currently, which limit the maximum drifts within the pre-selected target drift
usually starts with an initial design according to conventional level (1.75%), as well as achieve the intended yield mechanism
elastic design procedures using applicable design codes, followed under the DBE hazard level. The maximum drifts are generally
by cumbersome and time-consuming iterative assessment process uniformly distributed along the building height.
by using inelastic static or dynamic analyses till the desired 2. Mean values of maximum story drift ratios of the study BRBFs
performance objectives are achieved. The iterations are carried under the MCE hazard level are approximately 4%. However,
out in a purely trial-and-error manner. No guidance is provided
previous experimental studies have shown that a well-detailed
to the designer as to how to achieve the desired goals such as,
BRBF will experience minor damage at this drift level.
controlling drifts, distribution and extent of inelastic deformation,
etc. In contrast, the PBPD method is a direct design method, which 3. No iteration is required to achieve the desired performance
generally requires no evaluation after the initial design because objectives of all BRBFs since the inelastic characteristics of
the nonlinear behavior and key performance criteria are built structural components and target drift ratio are directly
into the design process from the beginning. In this study, this considered in the design. Further, due to well-controlled drift,
design methodology is applied to three low-to-medium rise BRBFs it is possible to achieve the desired seismic performance by
and the robustness and versatility of this method was evaluated neglecting additional force due to P-Delta effect in the design
by the seismic performance of these BRBFs under forty recorded for simplicity.
ground motions representing the DBE and the MCE hazard levels. 4. The PBPD design base shears for the 3-, 6-, and 9-story BRBFs are
Following conclusions are drawn from the present study: 91%, 57%, and 64% of that calculated based on the modern codes.
Author's personal copy

2958 D.R. Sahoo, S.-H. Chao / Engineering Structures 32 (2010) 2950–2958

This indicates that a more economical design can be realized [14] Goel SC, Chao SH. Performance-based plastic design—earthquake resistant
by the PBPD method, while maintaining the desirable seismic steel structures. International Code Council; 2008. p. 261.
[15] Newmark NM, Hall WJ. Earthquake spectra and design. El Cerrito (CA):
Earthquake Engineering Research Institute; 1982.
[16] ASCE 7-05. Minimum design load for buildings and other structures. Reston
References (VA): American Society of Civil Engineers; 2005.
[17] Sahoo DR, Chao SH. Performance-based plastic design for buckling-restrained
[1] López WA, Sabelli R. Seismic design of buckling-restrained braced frames. Steel braced frames. In: Proceedings of 9th US national and 10th canadian
tips. Moraga (CA): Structural Steel Educational Council; 2004. conference on earthquake engineering. 2010 [in press].
[2] Huang YC, Tsai KC. Experimental responses of large scale buckling restrained [18] Chao SH, Goel SC, Lee SS. A seismic design lateral force distribution based on
brace frames. Report no. CEER/R91-03. National Taiwan University; 2002 inelastic state of structures. Earthq Spectra 2007;23(5):547–69.
[in Chinese]. [19] Merritt S, Uang CM, Benzoni G. Subassemblage testing of star seismic buckling-
[3] Richard RM. Braced frame steel structures 402: when and why frame action restrained braces. TR-2003/04. La Jolla (CA): Univ. of California at San Diego;
matters. Struct Eng 2009. 2003.
[4] Uang CM, Nakashima M. Steel buckling-restrained braced frames. In: Earth- [20] AISC. ANSI/AISC 341-05. Seismic provisions for structural steel buildings.
quake engineering: recent advances and applications. Boca Raton (FL): CRC Chicago (IL): American Institute of Steel Construction; 2005.
Press; 2000. [21] Gupta A, Krawinkler H. Prediction of seismic demands for SMRFs with ductile
[5] Sabelli R. Research on improving the design and analysis of earthquake
connections and elements. Report no. SAC/BD-99/06. Sacramento (CA): SAC
resistant steel braced frames. Final report—NEHRP Fellowship in Earthquake
Joint Venture; 1999.
Hazard Reduction; 2000.
[6] Fahnestock LA, Ricles JM, Sause R. Experimental evaluation of a large-scale [22] SEAOC. Recommended lateral force requirements and commentary. 7th ed.
buckling-restrained braced frame. J Struct Eng 2007;133(9):1205–14. Sacramento (CA): Seismology Committee of Structural Engineers Association
[7] Prasad BK. Current status of buckling-restrained braced frame design: of California; 1999.
currently available buckling-restrained braces. In: Proceedings of 72nd annual [23] Building Seismic Safety Council. BSSC. National earthquake hazard reduction
convention. SEAOC. 2004. program (NEHRP) recommended provisions for seismic regulations for
[8] Federal Emergency Management Agency. FEMA. Next-generation new buildings and other structures—part 2: commentary. FEMA 450-1.
performance-based seismic design guidelines—program plan for new and Washington (DC): Federal Emergency Management Agency; 2003.
existing buildings. FEMA-445. Washington (DC); 2006. [24] AISC. ANSI/AISC 360-05. Specifications for structural steel buildings. Chicago
[9] Housner G. Limit design of structures to resist earthquakes. In: Proceedings of (IL): American Institute of Steel Construction; 2005.
the first world conference on earthquake engineering. 1956. [25] Richards PW. Seismic column demand in ductile braced frames. J Struct Eng
[10] Lee SS, Goel SC. Performance-based design of steel moment frames using
target drift and yield mechanism. Report no. UMCEE 01-17; MI: University of
[26] Thornton W, Muir LS. Design of vertical bracing connections for high-seismic
Michigan at Ann Arbor; 2001.
[11] Chao SH, Goel SC. Performance-based design of eccentrically braced frames drift. Modern Steel Construction; March 2009.
using target drift and yield mechanism. AISC Eng J 2006;173–200. 3rd Quarter. [27] Tremblay R, Bolduc P, Neville R, DeVall R. Seismic testing and performance of
[12] Chao SH, Goel SC. Performance-based plastic design of special truss moment buckling restrained bracing systems. Canad J Civ Eng 2006;33:183–98.
frames. AISC Eng J 2008;127–50. 2nd Quarter. [28] Somerville PG, Smith M, Punyamurthula S, Sun J. Development of ground
[13] CSI. PERFORM 3D user’s manual-version 4.0. Berkeley, (CA): Computers and motion time histories for phase 2 of the FEMA/SAC steel project. Report no.
Structures Inc.; 2007. SAC/BD-97/04. Sacramento (CA): SAC Joint Venture; 1997.